ConsID-Gen: View-Consistent and Identity-Preserving Image-to-Video Generation

TL;DR

ConsID-Gen introduces a view-assisted image-to-video generation framework that enhances multi-view consistency and identity preservation by leveraging auxiliary views and a dual-stream encoder, outperforming existing models.

Contribution

The paper presents a new dataset, ConsIDVid, and a benchmarking framework, along with a novel view-assisted generation method, improving identity fidelity and temporal coherence in image-to-video tasks.

Findings

ConsID-Gen outperforms state-of-the-art models on ConsIDVid-Bench.

The framework achieves superior identity preservation and temporal coherence.

Experiments validate the effectiveness of auxiliary views and dual-stream encoding.

Abstract

Image-to-Video generation (I2V) animates a static image into a temporally coherent video sequence following textual instructions, yet preserving fine-grained object identity under changing viewpoints remains a persistent challenge. Unlike text-to-video models, existing I2V pipelines often suffer from appearance drift and geometric distortion, artifacts we attribute to the sparsity of single-view 2D observations and weak cross-modal alignment. Here we address this problem from both data and model perspectives. First, we curate ConsIDVid, a large-scale object-centric dataset built with a scalable pipeline for high-quality, temporally aligned videos, and establish ConsIDVid-Bench, where we present a novel benchmarking and evaluation framework for multi-view consistency using metrics sensitive to subtle geometric and appearance deviations. We further propose ConsID-Gen, a view-assisted I2V generation framework that augments the first frame with unposed auxiliary views and fuses semantic and structural cues via a dual-stream visual-geometric encoder as well as a text-visual connector, yielding unified conditioning for a Diffusion Transformer backbone. Experiments across ConsIDVid-Bench demonstrate that ConsID-Gen consistently outperforms in multiple metrics, with the best overall performance surpassing leading video generation models like Wan2.1 and HunyuanVideo, delivering superior identity fidelity and temporal coherence under challenging real-world scenarios. We will release our model and dataset at https://myangwu.github.io/ConsID-Gen.